About this Role
We're building a forward-deployed engineering practice at the frontier of AI-native software development and we're looking for someone who wants to be on the ground floor of that. This is not a consulting role where you deliver recommendations from the outside. You'll embed as a technical insider within enterprise engineering organizations, work alongside their teams daily, and be directly responsible for shipping production AI agents, driving tool adoption, and raising the AI fluency of every engineer you work with.
Responsibilities
- Join the client's daily engineering rituals: standups, code reviews, sprint planning, and incident response
- Serve as the primary technical authority for AI-assisted and agentic development within assigned squads
- Define coding standards, prompt engineering playbooks, and guardrails teams can follow independently
- Bridge engineering managers, product owners, and security stakeholders with hands-on implementation
- Design and ship autonomous agents using Claude API, Claude Code, OpenAI Codex, or equivalent platforms
- Build multi-step agent workflows: task decomposition, tool calling, subagent orchestration, error handling, and rollback
- Deploy agents that run in parallel with existing teams — accelerating delivery without disrupting live workflows
- Create evaluation harnesses (golden datasets, regression suites, LLM-as-judge) to ensure agent reliability in production
- Structure API documentation, architecture patterns, and domain knowledge into machine-readable context for AI tools
- Implement layered context strategies: rules files (CLAUDE.md / AGENTS.md), subdirectory conventions, skills, hooks, and MCP integrations
- Apply the right retrieval approach per workload: agentic just-in-time search for live code; RAG/vector indexing for stable documentation
- Onboard engineering teams to Claude Code, Cursor, Codex, and related tools with team-specific guardrails
